\BOOKMARK [1][-]{section.1}{Introduction}{}% 1
\BOOKMARK [1][-]{section.2}{Background}{}% 2
\BOOKMARK [2][-]{subsection.2.1}{Methane emissions, detection, and repair}{section.2}% 3
\BOOKMARK [2][-]{subsection.2.2}{Regulatory setting}{section.2}% 4
\BOOKMARK [1][-]{section.3}{Model}{}% 5
\BOOKMARK [1][-]{section.4}{Data description}{}% 6
\BOOKMARK [1][-]{section.5}{Estimating treatment effects}{}% 7
\BOOKMARK [2][-]{subsection.5.1}{Full sample}{section.5}% 8
\BOOKMARK [2][-]{subsection.5.2}{Sites with zero baseline leakage}{section.5}% 9
\BOOKMARK [1][-]{section.6}{Discussion}{}% 10
\BOOKMARK [2][-]{subsection.6.1}{Can conditional information provision improve outcomes?}{section.6}% 11
\BOOKMARK [2][-]{subsection.6.2}{Monitoring, efficiency, and technical constraints}{section.6}% 12
\BOOKMARK [1][-]{section.7}{Conclusion}{}% 13
\BOOKMARK [1][-]{appendix.A}{Proofs for section 3}{}% 14
\BOOKMARK [1][-]{appendix.B}{Empirical Appendix}{}% 15
\BOOKMARK [2][-]{subsection.B.1}{Additional Data Background}{appendix.B}% 16
\BOOKMARK [2][-]{subsection.B.2}{Operator descriptives and testing SUTVA}{appendix.B}% 17
\BOOKMARK [2][-]{subsection.B.3}{Balance using CDFs}{appendix.B}% 18
\BOOKMARK [2][-]{subsection.B.4}{Power, type S errors, and exaggeration ratios}{appendix.B}% 19
\BOOKMARK [2][-]{subsection.B.5}{Sites with zero baseline leakage}{appendix.B}% 20
\BOOKMARK [3][-]{subsubsection.B.5.1}{Treatment intensity}{subsection.B.5}% 21
\BOOKMARK [2][-]{subsection.B.6}{Sites with above-median baseline leakage}{appendix.B}% 22
\BOOKMARK [2][-]{subsection.B.7}{Sites with non-zero and below-median baseline leakage}{appendix.B}% 23
\BOOKMARK [2][-]{subsection.B.8}{Repair status}{appendix.B}% 24
